JP2015123899A - Vehicle surrounding-situation estimation apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a vehicle surrounding-situation estimation apparatus capable of appropriately estimating a surrounding situation of a vehicle even in a case where abnormality occurs to a sensor due to a collision, disabling detection of a surrounding situation.SOLUTION: A vehicle surrounding-situation estimation apparatus includes: collision detection means for detecting an event of a vehicle itself colliding with an object outside of the vehicle; surrounding-situation detection means for detecting a situation of a detection zone surrounding the vehicle itself; surrounding-situation estimation means for estimating a situation of a prediction zone surrounding the vehicle itself when the collision occurs on the basis of the detection result of the surrounding-situation detection means before the collision detection means detects the collision; surrounding-situation record means for recording a situation of the prediction zone estimated by the surrounding-situation estimation means; and surrounding-situation prediction means for predicting a situation of a vehicle-surrounding zone that is surrounding the vehicle itself, after the collision detection means detects the collision, on the basis of a situation of the prediction zone recorded by the surrounding-situation record means before the collision detection means detects the collision.

Description

  The present invention relates to a vehicle surrounding situation estimation device.

  Conventionally, after detecting a primary accident (primary collision) due to a collision between a vehicle and an obstacle outside the vehicle, there is a technique for reducing damage caused by a secondary accident (secondary collision) that may occur subsequently. It has been reported. For example, Patent Document 1 discloses a technique for detecting and analyzing a surrounding environment of a vehicle after a collision by a sensor and performing traveling control such as a brake operation or a steering operation according to a comparison result between the surrounding environment and the behavior of the vehicle. ing.

Special table 2007-516127

  By the way, in the prior art (Patent Document 1 or the like), when an abnormality such as sensor breakage occurs due to a primary accident, it becomes impossible to detect the surrounding environment of the vehicle. As a result, in the related art, it may be difficult to estimate the surrounding situation of the vehicle based on the detection result of the surrounding environment and to control the traveling of the vehicle.

  The present invention has been made in view of the above circumstances, and it is possible to suitably estimate the surrounding situation of the vehicle even when the sensor is abnormal due to a collision and the surrounding situation cannot be detected. An object of the present invention is to provide a vehicle surrounding situation estimation device.

  The vehicle surrounding state estimation device of the present invention includes a collision detecting unit that detects that the own vehicle has collided with an object outside the vehicle, a surrounding state detecting unit that detects a state of a detection region around the own vehicle, and the collision detecting unit. Based on the detection result of the surrounding situation detection means before the collision is detected, the surrounding situation estimation means for estimating the situation of the prediction area around the host vehicle at the time of collision, and the surrounding situation estimation means estimated by the surrounding situation estimation means A surrounding situation recording means for recording the situation of the prediction area; and after the collision detection means detects a collision, the situation of the prediction area recorded in the surrounding situation recording means before the collision detection means detects a collision. On the basis of this, a surrounding situation prediction means for predicting the situation of the vehicle surrounding area around the host vehicle is provided.

  The vehicle surrounding state estimation device further includes collision avoidance determining means for determining whether or not a collision between the host vehicle and an object outside the vehicle detected by the surrounding state detecting unit can be avoided, and the surrounding state estimation The means estimates the situation of the prediction area after the collision avoidance determining means determines that the collision avoidance is impossible, and the surrounding situation recording means determines the time point after the collision avoidance determining means determines that the collision avoidance is impossible It is preferable to record the state of the prediction region estimated by the surrounding state estimation unit at a time from when the collision detection unit detects a collision.

In the surrounding vehicle state estimation device, after the collision detecting unit detects a collision, a sensor abnormality determining unit that determines presence / absence of abnormality in the plurality of surrounding state detecting units mounted on the host vehicle,
The surrounding situation predicting means further comprises a detection region set before the collision of the surrounding situation detecting means determined to be abnormal by the sensor abnormality judging means after the collision detecting means detects a collision. The situation of the corresponding abnormality recognition area in the vehicle surrounding area is predicted based on the situation of the prediction area recorded in the surrounding situation recording means before the collision detecting means detects a collision, and the sensor abnormality The situation of the normal recognition area in the vehicle surrounding area corresponding to the detection area of the surrounding situation detecting means determined to be normal by the judging means is predicted based on the detection result of the surrounding situation detecting means, It is preferable to predict the situation around the host vehicle.

  In the vehicle surrounding situation estimation apparatus, after the collision detecting unit detects a collision, traveling is performed to control the behavior of the host vehicle based on the situation of the vehicle surrounding area predicted by the surrounding situation predicting unit. Preferably, the control means is further provided.

  The vehicle surrounding situation estimation device according to the present invention has an effect that the surrounding situation of the vehicle can be suitably estimated even when an abnormality occurs in the sensor due to a collision and the surrounding situation cannot be detected.

FIG. 1 is a diagram showing a configuration of a vehicle surrounding situation estimation apparatus according to the present invention. FIG. 2 is a diagram illustrating an example of detection areas of a plurality of surrounding environment recognition sensors mounted on a vehicle. FIG. 3 is a diagram illustrating an example of predicting the situation of the vehicle surrounding area around the own vehicle based on the detection result of the surrounding environment recognition sensor. FIG. 4 is a diagram illustrating an example of a primary collision scene. FIG. 5 is a diagram illustrating an example of estimating a situation of a prediction region in which occurrence of a secondary collision around the host vehicle at the time of the collision is predicted based on the surrounding environment information acquired before the collision is detected. . FIG. 6 is a diagram illustrating an example of a situation in which the degree of coincidence of surrounding environment information is confirmed in an overlapping area between sensors. FIG. 7 is a diagram illustrating an example of a situation in which the degree of coincidence of the surrounding environment information is confirmed in the overlapping area between the sensors immediately before the collision. FIG. 8 is a diagram illustrating an example of a situation in which it is determined normal in the sensor abnormality determination performed immediately after the collision. FIG. 9 is a diagram illustrating an example of a situation where an abnormality is determined in the sensor abnormality determination performed immediately after the collision. FIG. 10 is a diagram illustrating an outline of the prediction process of the secondary collision occurrence area in the present embodiment. FIG. 11 is a flowchart showing an example of basic processing of the vehicle surrounding situation estimation apparatus according to the present invention. FIG. 12 is a flowchart illustrating an example of the surrounding situation estimation process immediately before the collision. FIG. 13 is a flowchart showing an example of the coincidence degree recording process immediately before the collision. FIG. 14 is a flowchart illustrating an example of sensor abnormality determination processing immediately after a collision. FIG. 15 is a flowchart illustrating an example of the surrounding situation prediction process according to the sensor state. FIG. 16 is a diagram illustrating an example of a scene in which a secondary collision position with respect to a moving object on a road is predicted. FIG. 17 is a diagram illustrating an example of a scene in which a secondary collision position with respect to a stationary object on a road is predicted.

  DESCRIPTION OF EMBODIMENTS Embodiments of a vehicle surrounding situation estimation apparatus according to the present invention will be described below in detail with reference to the drawings. In addition, this invention is not limited by this embodiment. In addition, constituent elements in the following embodiments include those that can be easily assumed by those skilled in the art or those that are substantially the same.

[Embodiment]
With reference to FIGS. 1-10, the structure of the vehicle surrounding condition estimation apparatus which concerns on this invention is demonstrated. FIG. 1 is a diagram showing a configuration of a vehicle surrounding situation estimation apparatus according to the present invention. FIG. 2 is a diagram illustrating an example of detection areas of a plurality of surrounding environment recognition sensors mounted on a vehicle. FIG. 3 is a diagram illustrating an example of predicting the situation of the vehicle surrounding area around the own vehicle based on the detection result of the surrounding environment recognition sensor. FIG. 4 is a diagram illustrating an example of a primary collision scene. FIG. 5 is a diagram illustrating an example of estimating a situation of a prediction region in which occurrence of a secondary collision around the host vehicle at the time of the collision is predicted based on the surrounding environment information acquired before the collision is detected. . FIG. 6 is a diagram illustrating an example of a situation in which the degree of coincidence of surrounding environment information is confirmed in an overlapping area between sensors. FIG. 7 is a diagram illustrating an example of a situation in which the degree of coincidence of the surrounding environment information is confirmed in the overlapping area between the sensors immediately before the collision. FIG. 8 is a diagram illustrating an example of a situation in which it is determined normal in the sensor abnormality determination performed immediately after the collision. FIG. 9 is a diagram illustrating an example of a situation where an abnormality is determined in the sensor abnormality determination performed immediately after the collision. FIG. 10 is a diagram illustrating an outline of the prediction process of the secondary collision occurrence area in the present embodiment.

  In this embodiment, ECU1 has a function as a vehicle surrounding condition estimation apparatus which estimates the condition around a vehicle based on the detection result of the surrounding environment recognition sensor 3 mounted in the vehicle. The ECU 1 also has a function as a sensor abnormality detection device that detects an abnormality of the surrounding environment recognition sensor 3 and a vehicle control device that performs driving support control for controlling the behavior of the vehicle. The ECU 1 is electrically connected to the vehicle momentum detection sensor 2, the surrounding environment recognition sensor 3, and the actuator 4. The ECU 1 performs arithmetic processing based on various signals input from the vehicle momentum detection sensor 2 and the surrounding environment recognition sensor 3. For example, the ECU 1 determines the presence / absence of a collision based on various signals, considers abnormalities in the surrounding environment recognition sensor 3 that may occur due to the collision, and uses the situation around the vehicle estimated immediately before the collision, Arithmetic processing such as estimating the situation around the vehicle immediately after the collision is performed. The ECU 1 also performs arithmetic processing such as determining the presence or absence of a collision based on various signals and determining the presence or absence of an abnormality in the surrounding environment recognition sensor 3 that may occur due to the collision. Further, the ECU 1 outputs a control signal based on these calculation processing results to the actuator 4 and performs driving support control for controlling the behavior of the vehicle by operating the actuator 4.

  The vehicle momentum detection sensor 2 is a vehicle momentum detection device that detects various information indicating the vehicle momentum. In the present embodiment, the vehicle momentum detection sensor 2 includes an acceleration sensor 2a, a yaw rate sensor 2b, and a vehicle speed sensor 2c.

  The acceleration sensor 2a is an acceleration detection device that detects acceleration applied to the vehicle body. The acceleration sensor 2a outputs an acceleration signal indicating the detected acceleration to the ECU 1.

  The yaw rate sensor 2b is a yaw rate detection device that detects the yaw rate of the vehicle. The yaw rate sensor 2b outputs a yaw rate signal indicating the detected yaw rate to the ECU 1.

  The vehicle speed sensor 2c is a wheel speed detection device that is provided for each wheel and detects each wheel speed. Each vehicle speed sensor 2c detects a wheel speed that is a rotational speed of each wheel. Each vehicle speed sensor 2c outputs a wheel speed signal indicating the detected wheel speed of each wheel to the ECU 1. The ECU 1 calculates the vehicle speed, which is the traveling speed of the vehicle, based on the wheel speed of each wheel input from each vehicle speed sensor 2c. The ECU 1 may calculate the vehicle speed based on the wheel speed input from at least one of the vehicle speed sensors 2c.

  As described above, the vehicle momentum detection sensor 2 detects the acceleration detected by the acceleration sensor 2a, the yaw rate detected by the yaw rate sensor 2b, and the wheel speed detected by the vehicle speed sensor 2c as various information indicating the vehicle momentum, These pieces of information are output to the ECU 1.

  The surrounding environment recognition sensor 3 is a surrounding environment recognition device that recognizes vehicle surrounding conditions such as moving objects and stationary obstacles around the vehicle. That is, the surrounding environment recognition sensor 3 functions as a surrounding state detection unit that detects the state of the detection region around the host vehicle. The surrounding environment recognition sensor 3 is configured by a radar, a camera, or the like. The surrounding environment recognition sensor 3 is, for example, as the surrounding environment information, for example, a relative position with respect to a guardrail or a white line on a road, a relative position of a surrounding stationary target, a surrounding moving target (for example, forward, backward, or sideward of a vehicle). Information such as a relative position, a relative speed, and a relative acceleration with respect to an existing moving target is acquired, and the surrounding environment information is output to the ECU 1. Furthermore, the surrounding environment recognition sensor 3 also includes information on the attributes of surrounding obstacles such as the strength, brightness, and color of the recognition target as the surrounding environment information in addition to information on the relative position and relative speed of the recognition target around the vehicle. You may acquire and output to ECU1. For example, when the surrounding environment recognition sensor 3 is configured by a radar, the wavelength pattern of the reflected wave of the radar differs depending on whether the intensity of an object to be recognized by the surrounding environment recognition sensor 3 is hard or soft. The surrounding environment recognition sensor 3 detects the intensity of the recognition target using the difference in wavelength pattern. The brightness and color of the recognition target are detected by the difference in the wavelength pattern of the reflected wave of the radar when the surrounding environment recognition sensor 3 is configured by a radar. Detected by difference in contrast.

  In the present embodiment, a plurality of surrounding environment recognition sensors 3 are mounted on the vehicle. For example, the surrounding environment recognition sensor 3 includes a sensor 1 as a first sensor, a sensor 2 as a second sensor, and a sensor 3 as a third sensor, as shown in FIG. The number of surrounding environment recognition sensors mounted on the vehicle is not limited to three as in the example of FIG. 1, and three or more sensors may be mounted on the vehicle.

  Sensors 1 to 3 detect the situation of different detection areas. For example, the sensor 1 functions as a first sensor that detects the state of the first region around the host vehicle. The sensor 2 functions as a second sensor that detects a situation of the second region around the host vehicle that is different from the first region and overlaps a part of the first region. The sensor 3 functions as a third sensor that detects the situation of the third region around the host vehicle that is different from the first region and the second region and overlaps a part of the first region.

  As an example, a case where the sensors 1 to 3 are attached to the front surface of the vehicle 10 as illustrated in FIG. 2 will be described as an example. In FIG. 2, the sensor 1 detects the state of a detection region (the first region in FIG. 2) that covers the traveling direction side of the vehicle 10. The sensor 2 detects the state of a detection region (second region in FIG. 2) that covers the right side from the right front side of the vehicle. The sensor 3 detects the state of a detection region (third region in FIG. 2) that covers the left side from the left front of the vehicle. The first area detected by the sensor 1 and the second area detected by the sensor 2 partially overlap. The straddle region between the sensor 1 and the sensor 2 is defined as a first overlap region. Further, the first area detected by the sensor 1 and the third area detected by the sensor 3 partially overlap. A region between the detection regions of the sensors 1 and 3 is defined as a second overlapping region. The sensor mounting position is not limited to the front surface as in the example of FIG. 2, and may be the right side surface, the left side surface, the rear surface, etc. in addition to the front surface of the vehicle.

  Returning to FIG. 1, the description of the configuration of the vehicle surrounding state estimation device according to the present invention will be continued. The ECU 1 includes a surrounding environment information acquisition unit 1a, a surrounding situation prediction unit 1b, a surrounding situation estimation unit 1c, a surrounding situation recording unit 1d, a coincidence degree recording unit 1e, a collision avoidance determination unit 1f, and a collision detection unit 1g. And at least a sensor abnormality determination unit 1h and a travel control unit 1i.

  In the ECU 1, the surrounding environment information acquisition unit 1a receives and acquires the surrounding environment information transmitted from the surrounding environment recognition sensor 3 and indicating the surrounding environment information such as a moving object and a stationary obstacle around the vehicle. It is an acquisition means. That is, the surrounding environment information acquisition unit 1a acquires the detection result of the surrounding environment recognition sensor 3 as the surrounding state detection means as the surrounding environment information. In the present embodiment, the surrounding environment information acquisition unit 1a includes, for example, a first area, a second area, and a third area that are transmitted from the sensors 3a, 3b, and 3c mounted as the surrounding environment recognition sensor 3, respectively. Receive and acquire ambient environment information indicating the status of

  The surrounding situation prediction unit 1b predicts the situation around the host vehicle based on the surrounding environment information acquired by the surrounding environment information acquisition unit 1a, that is, based on the detection result of the surrounding environment recognition sensor 3 as the surrounding state detection unit. It is a situation prediction means. For example, the surrounding situation prediction unit 1b predicts the situation of the vehicle surrounding area around the host vehicle. The vehicle surrounding area is an area around the current position of the host vehicle that is different from the detection area of the surrounding environment recognition sensor 3, and is an area that is set to a range that includes the detection areas of the plurality of surrounding environment recognition sensors 3. . For example, the surrounding situation prediction unit 1b uses surrounding environment information indicating the relative position, relative speed, attribute, and the like of an object outside the vehicle detected in the detection area of the plurality of surrounding environment recognition sensors 3 mounted on the vehicle. Thus, the situation of an obstacle that may collide with the vehicle in the vehicle surrounding area is predicted.

  As an example, as shown in FIG. 3, sensors 1 to 3 are attached to the front surface of the vehicle 10, the sensor 4 is the right side surface of the vehicle 10, the sensor 5 is the left side surface of the vehicle 10, and the sensor 6 is the rear surface of the vehicle 10. The case where it is attached will be described as an example. In FIG. 3, sensors 1 to 6 detect the situation of different detection areas. The sensor 1 is a first area on the traveling direction side, the sensor 2 is a second area on the right side from the right front, the sensor 3 is a third area on the left side from the left front, and the sensor 4 is a third area on the right side from the right rear. The four areas, the sensor 5 detect the situation of the fifth area on the left side from the left rear, and the sensor 6 detects the situation of the sixth area on the side opposite to the traveling direction. And the surrounding situation prediction part 1b uses the surrounding environment information which shows the condition in the 1st area | region-6th area | region of each sensor 1-6 acquired by the surrounding environment information acquisition part 1a, and the vehicle 10 is presently located. The situation of the vehicle surrounding area set to a predetermined range around the vehicle position is predicted.

  In the present embodiment, the surrounding situation prediction unit 1b predicts the situation of the vehicle surrounding area by specifying the current position of an object outside the vehicle based on the surrounding environment information transmitted from the surrounding environment information acquisition unit 1a. For example, the surrounding situation prediction unit 1b specifies the current position of the obstacle around the host vehicle based on the detection results of the plurality of surrounding environment recognition sensors 3. Specifically, the surrounding situation prediction unit 1b specifies the current position of the obstacle based on the relative position, relative speed, relative acceleration, and the like between the obstacle and the host vehicle included in the surrounding environment information.

  Here, in addition to the surrounding environment information acquired by the surrounding environment information acquisition unit 1a, the surrounding situation prediction unit 1b further uses various information indicating the vehicle momentum transmitted from the vehicle momentum detection sensor 2 to use the position of the host vehicle. In addition, the situation around the host vehicle may be predicted after specifying the state of inclination.

  Returning to FIG. 1, the description of the configuration of the vehicle surrounding state estimation device according to the present invention will be continued. Of the ECU 1, the surrounding situation estimation unit 1 c is based on the surrounding environment information acquired by the surrounding environment information acquisition unit 1 a before the collision detection unit 1 g detects a collision, It is a surrounding situation estimation means for estimating the situation.

  For example, as shown in FIG. 4, a scene in which the vehicle 10 as the host vehicle overtakes another vehicle 30 that precedes the same traveling lane and first collides with another vehicle 20 traveling in the opposite lane is This represents a situation in which the occurrence of a secondary collision between the vehicle 10 and the vehicle 30 is predicted. In such a situation, as shown in the lower diagram of FIG. 5, the surrounding situation estimation unit 1c determines that the collision avoidance determination unit 1f cannot avoid the collision with the vehicle 20, and then the collision detection unit 1g detects that the vehicle 20 As shown in the upper diagram of FIG. 5 based on the detection result of the surrounding environment recognition sensor 3 before detecting a collision with the vehicle (in FIG. In addition, the situation of the prediction area around the host vehicle when it collides with the vehicle 20 is estimated. In the present embodiment, the prediction area is an area set in a predetermined range around a prediction position where the vehicle 10 primarily collides with an object outside the vehicle (vehicle 20 in FIG. 5) after a predetermined time has elapsed. The range of the prediction area may be the same as or different from the range of the vehicle surrounding area.

  In this embodiment, the surrounding state estimation unit 1c predicts the state of the prediction region by predicting the position of an object outside the vehicle after a lapse of a predetermined time based on the surrounding environment information transmitted from the surrounding environment information acquisition unit 1a. presume. The predetermined time is set based on the time until the collision calculated by the collision avoidance determination unit 1f. For example, the surrounding state estimation unit 1c predicts the position of an obstacle around the host vehicle after a predetermined time has elapsed based on the detection results of the plurality of surrounding environment recognition sensors 3. The surrounding state estimation unit 1c predicts the moving position after a predetermined time as the position of the obstacle based on the relative position, relative speed, relative acceleration, and the like between the obstacle and the vehicle included in the surrounding environment information. For example, the predetermined time is set to a time required for the vehicle 10 to collide with an obstacle.

  Here, in addition to the surrounding environment information acquired by the surrounding environment information acquisition unit 1a, the surrounding state estimation unit 1c further uses various information indicating the vehicle momentum transmitted from the vehicle momentum detection sensor 2 to perform the collision. You may estimate the condition of the prediction area | region where generation | occurrence | production of the secondary collision around the own vehicle at the time of a collision is estimated after estimating the position and inclination of the own vehicle.

  Returning to FIG. 1, the description of the configuration of the vehicle surrounding state estimation device according to the present invention will be continued. In the ECU 1, the surrounding situation recording unit 1 d is a surrounding situation recording unit that records the situation of the prediction area around the host vehicle at the time of the collision estimated by the surrounding situation estimation unit 1 c. In the present embodiment, the surrounding situation recording unit 1d is estimated by the surrounding situation estimating unit 1c from the time after the collision avoidance determining unit 1f determines that the collision avoidance is impossible until the collision detecting unit 1g detects the collision. Record the status of the predicted region. For example, the surrounding situation recording unit 1d transmits the situation of the prediction region estimated by the surrounding situation estimation unit 1c in association with the estimated time in the memory of the ECU 1 and records it.

  The coincidence degree recording unit 1e is a coincidence degree recording unit that calculates and records the coincidence degree of the surrounding environment information in the overlapping region of the surrounding environment recognition sensor 3 based on the surrounding environment information acquired by the surrounding environment information acquisition unit 1a. is there.

  As an example, as shown in FIG. 6, the coincidence degree recording unit 1e calculates and records the coincidence degree of the surrounding environment information in the straddle region between the sensors. In FIG. 6, the coincidence degree recording unit 1 e is located on the right side (the position (i) in FIG. 6) of the vehicle 10 to be recognized in the overlapping region (the first overlapping region in FIG. 6) of the sensor 1 and the sensor 2. ), The degree of coincidence is calculated and recorded.

  In such a case, for example, the coincidence degree recording unit 1e obtains the recognition target wall detected from the surrounding environment information acquisition unit 1a as the surrounding environment information of the sensor 1 in the first overlapping region in the first region. Information including the relative position, the strength indicating the hardness and softness of the wall itself, the brightness of the wall, the color of the wall, etc. is received. Further, the coincidence degree recording unit 1e receives, as the surrounding environment information of the sensor 2 from the surrounding environment information acquiring unit 1a, the relative position of the wall to be recognized detected in the first overlapping region in the second region, the wall Receive information including intensity, brightness, color, etc. Then, the coincidence degree recording unit 1e compares the surrounding environment information of the sensor 1 and the surrounding environment information of the sensor 2 for each parameter (relative position with respect to the wall, intensity, brightness, and color in FIG. 6). Subsequently, in the coincidence degree recording unit 1e, the parameters to be compared are the same between the sensor 1 and the sensor 2, or are different between the sensor 1 and the sensor 2, but the difference is within a predetermined threshold range. If it is within, it is determined that the matching degree is high. For example, when comparing the relative position with the wall detected by the sensor 1 and the relative position with the wall detected by the sensor 2, the coincidence degree recording unit 1 e includes the mounting position of any one of the sensor 1 and the sensor 2, Alternatively, a predetermined position on the vehicle is set as the reference position. Then, the coincidence degree recording unit 1e calculates the relative position between the reference position and the wall, and compares the calculated relative positions to determine the coincidence. For information on wall attributes (for example, wall strength, brightness, color, etc.), the degree of coincidence is determined by comparing the conditions detected by sensor 1 and sensor 2 respectively.

  On the other hand, the coincidence degree recording unit 1e has a parameter to be compared (for example, a relative position between the wall detected by the sensor 1 and a relative position between the wall detected by the sensor 2) between the sensor 1 and the sensor 2. If the difference is outside the predetermined threshold range, it is determined that the degree of coincidence is low. In addition, the coincidence degree recording unit 1e determines that the degree of coincidence of the surrounding environment information is low even when there is no overlapping area between the sensor 1 and the sensor 2. Then, the coincidence degree recording unit 1e performs a process for determining the degree of coincidence of the surrounding environment information between the sensor 1 and the sensor 2 for each parameter to be compared (for example, for each wall strength, brightness, and color). And the degree of coincidence of the surrounding environment information is calculated based on the degree of coincidence determined for each parameter. The degree of coincidence of the surrounding environment information may be, for example, a score obtained by scoring the degree of coincidence determined for each parameter and totaling them.

  In the example of FIG. 6, for convenience of explanation, only the calculation of the degree of coincidence of the surrounding environment information related to the wall detected in the first overlapping area that is the straddling area between the sensor 1 and the sensor 2 has been described. The coincidence degree recording unit 1e calculates the coincidence degree of the surrounding environment information for each pair of sensors having overlapping regions. For example, the coincidence degree recording unit 1e calculates the degree of coincidence of the surrounding environment information regarding the wall detected in the first overlapping area, and also recognizes the recognition target detected in the second overlapping area in the first area of the sensor 1. And the surrounding environment information related to the recognition target detected in the second overlapping area in the third area of the sensor 3 are calculated for each parameter, and the degree of coincidence of the surrounding environment information is calculated. Then, the coincidence degree recording unit 1e transmits and records the calculated coincidence degree in the memory of the ECU 1 in association with the calculation time.

  In the present embodiment, the coincidence degree recording unit 1e calculates and records the coincidence degree of the surrounding environment information at a predetermined timing. For example, the coincidence degree recording unit 1e detects the timing immediately before the collision (that is, the timing when the collision avoidance determination unit 1f determines that the collision cannot be avoided) or the timing immediately after the collision (that is, the collision detection unit 1g detects the collision). The degree of coincidence of the surrounding environment information in the overlapping area of the surrounding environment recognition sensor 3 is calculated and recorded.

  Returning to FIG. 1, the description of the configuration of the vehicle surrounding state estimation device according to the present invention will be continued. Of the ECU 1, the collision avoidance determination unit 1 f is based on various information indicating the vehicle momentum transmitted from the vehicle momentum detection sensor 2 and the surrounding environment information transmitted from the surrounding environment information acquisition unit 1 a. It is a collision avoidance judging means for judging whether or not a collision with an external object can be avoided. For example, the collision avoidance determination unit 1f determines the vehicle position based on the relative position and relative speed between the object outside the vehicle indicated by the surrounding environment information and the vehicle 10, the vehicle speed and acceleration of the vehicle 10 included in the information indicated by the vehicle momentum, and the like. The time until the collision between the external object and the vehicle 10 (so-called collision prediction time (Time-To-Collision: TTC)) is calculated. Then, the collision avoidance determination unit 1f determines that collision can be avoided if the calculated TTC is equal to or greater than a predetermined threshold, and determines that collision avoidance is not possible if the calculated TTC is less than the predetermined threshold.

  The collision detection unit 1g causes the host vehicle to collide with an object outside the vehicle based on various information indicating the vehicle momentum transmitted from the vehicle momentum detection sensor 2 and the surrounding environment information transmitted from the surrounding environment information acquisition unit 1a. This is a collision detection means for detecting the occurrence. The collision detection unit 1g, for example, based on a change in the relative position between the collision target indicated by the surrounding environment information and the vehicle 10, the change in the acceleration and yaw rate of the vehicle 10 included in the information indicated by the vehicle momentum, and the like. To detect a collision.

  The sensor abnormality determination unit 1h is a sensor abnormality determination unit that determines whether there is an abnormality in the surrounding environment recognition sensors 3 mounted on the host vehicle after the collision detection unit 1g detects a collision. In the present embodiment, the sensor abnormality determination unit 1h is a first sensor that detects the situation of the first region around the host vehicle, and a region that is different from the first region and overlaps a part of the first region. The presence or absence of abnormality of the second sensor that detects the situation of the second region around the subject vehicle is determined. After the collision detection unit 1g detects a collision, the sensor abnormality determination unit 1h is normal in the situation where there is an overlapping region in which the first region and the second region partially overlap. In the situation where the first area and the second area do not overlap in the overlapping area, it is determined that at least one of the first sensor and the second sensor is abnormal. Specifically, after the collision detection unit 1g detects a collision, the sensor abnormality determination unit 1h determines that the first sensor and the second sensor do not detect the same situation in the overlapping region. It is determined that at least one of the sensor and the second sensor is abnormal.

  As an example, referring to FIG. 7 to FIG. 9, the first sensor is the sensor 1, the second sensor is the sensor 2, and the overlapping region where the first region and the second region partially overlap is the first A process in which the sensor abnormality determination unit 1h determines whether or not there is a sensor abnormality when the overlapping area is used will be described.

  FIG. 7 shows a situation where the vehicle 10 as the host vehicle cannot avoid a collision with another vehicle 20 as a moving object around the vehicle. In the situation as shown in FIG. 7, first, the collision avoidance determination unit 1f of the ECU 1 first determines the relative position and relative speed between the vehicle 20 and the vehicle 10 indicated by the surrounding environment information and the information included in the information indicated by the vehicle momentum. Based on the vehicle speed, acceleration, and the like, a time (TTC) until the collision between the vehicle 20 and the vehicle 10 is calculated. The collision avoidance determination unit 1f determines that collision avoidance is impossible because the calculated TTC is less than the predetermined threshold. Subsequently, the coincidence degree recording unit 1e of the ECU 1 matches the surrounding environment information in the overlapping region of the surrounding environment recognition sensor 3 when the collision avoidance determining unit 1f determines that the collision avoidance is impossible (that is, the timing immediately before the collision). Calculate and record the degree. Specifically, in the example of FIG. 7, the coincidence degree recording unit 1 e receives the vehicle 20 detected in the first overlapping region in the first region as the surrounding environment information of the sensor 1 from the surrounding environment information acquisition unit 1 a. Information including the relative position of the vehicle, the intensity of the vehicle 20, brightness, color, and the like. In addition, the coincidence degree recording unit 1e receives the relative position with respect to the vehicle 20 detected in the first overlapping region in the second region, the strength of the vehicle 20, as the surrounding environment information of the sensor 2 from the surrounding environment information acquiring unit 1a. Receive information including brightness, color, etc. Then, the coincidence degree recording unit 1e compares the surrounding environment information of the sensor 1 and the surrounding environment information of the sensor 2 for each parameter (relative position, intensity, brightness, and color with respect to the vehicle 20 in FIG. 7). The degree of coincidence is calculated, and the calculated degree of coincidence is transmitted and recorded in the memory of the ECU 1 in association with the calculation time. In the example of FIG. 7, the relative position of the vehicle 20 detected by the sensor 1 in the first overlap region, the intensity, brightness, and color of the vehicle 20 and the vehicle 20 detected by the sensor 2 in the first overlap region. Since the relative position, the intensity, the brightness, and the color of the vehicle 20 are approximately the same, the coincidence degree recording unit 1e relates to the vehicle 20 that is detected in the first overlap region that is the straddle region between the sensor 1 and the sensor 2. It is recorded that the degree of coincidence of the surrounding environment information is high.

  FIG. 8 shows a situation immediately after the vehicle 10 as the own vehicle collides with another vehicle 20 as a moving object around the vehicle, and the sensor is operating normally even by the collision. In the situation shown in FIG. 8, first, the collision detection unit 1g of the ECU 1 detects the relative position between the vehicle 20 and the vehicle 10 indicated by the surrounding environment information, the acceleration and yaw rate of the vehicle 10 included in the information indicated by the vehicle momentum, and the like. Based on this change, a collision between the vehicle 20 and the vehicle 10 is detected. Then, the coincidence degree recording unit 1e of the ECU 1 calculates the degree of coincidence of the surrounding environment information in the overlapping region of the surrounding environment recognition sensor 3 when a collision is detected by the collision detecting unit 1g (that is, the timing immediately after the collision). Record. Specifically, in the example of FIG. 8, the coincidence degree recording unit 1 e includes the surrounding environment information regarding the vehicle 20 acquired in the first overlapping area in the first area of the sensor 1 and the second area in the sensor 2. The degree of coincidence with the surrounding environment information related to the vehicle 20 acquired in the first overlapping area is calculated and recorded. In the example of FIG. 8, the relative position of the vehicle 20 detected by the sensor 1 in the first overlap region, the intensity, brightness, and color of the vehicle 20 and the vehicle 20 detected by the sensor 2 in the first overlap region. Since the relative position, the intensity, the brightness, and the color of the vehicle 20 are approximately the same, the coincidence degree recording unit 1e relates to the vehicle 20 that is detected in the first overlap region that is the straddle region between the sensor 1 and the sensor 2. It is recorded that the degree of coincidence of the surrounding environment information is high.

  Subsequently, as shown in the example of FIG. 7, the sensor abnormality determination unit 1 h of the ECU 1 loads the degree of coincidence of the surrounding environment information recorded by the degree of coincidence recording unit 1 e at the timing immediately before the collision from the memory of the ECU 1. Then, the sensor abnormality determination unit 1h includes the degree of coincidence of the surrounding environment information regarding the vehicle 20 at the timing immediately before the loaded collision and the vicinity recorded by the coincidence degree recording unit 1e at the timing immediately after the collision as shown in the example of FIG. Compare the degree of coincidence of environmental information. As a result of the comparison, the sensor abnormality determination unit 1h has a high degree of coincidence of the surrounding environment information at the timing immediately before the collision as shown in the example of FIG. 7, and the surrounding environment at the timing immediately after the collision as shown in the example of FIG. When the degree of coincidence of information is also high, since the degree of coincidence between the two is the same, it is determined that no abnormality has occurred in sensor 1 and sensor 2 before and after the collision. This is because no abnormality such as an axis deviation occurs in either sensor 1 or sensor 2 due to the collision, and neither the first area covered by sensor 1 nor the second area covered by sensor 2 changes due to the collision. Because it is thought that it was. Thus, after the collision detection unit 1g detects a collision, the sensor abnormality determination unit 1h includes the first sensor and the second sensor in a situation where there is an overlapping region where the first region and the second region partially overlap. This sensor is determined to be normal. Specifically, after the collision detection unit 1g detects a collision, the sensor abnormality determination unit 1h detects the first situation when the first sensor and the second sensor detect the same situation in the overlapping region. It is determined that the sensor and the second sensor are normal.

  FIG. 9 shows a situation immediately after the vehicle 10 as the own vehicle collides with another vehicle 20 as a moving object around the vehicle, and an abnormality occurs in the sensor due to the collision. In the situation shown in FIG. 9, first, the collision detection unit 1g of the ECU 1 detects the relative position between the vehicle 20 and the vehicle 10 indicated by the surrounding environment information, the acceleration and yaw rate of the vehicle 10 included in the information indicated by the vehicle momentum, and the like. Based on this change, a collision between the vehicle 20 and the vehicle 10 is detected. Then, the coincidence degree recording unit 1e of the ECU 1 calculates the degree of coincidence of the surrounding environment information in the overlapping region of the surrounding environment recognition sensor 3 when a collision is detected by the collision detecting unit 1g (that is, the timing immediately after the collision). Process to record. However, in the example of FIG. 9, the sensor 2 mounted on the vehicle 10 due to a collision with the vehicle 20 has an abnormality such as an axis deviation and the second region is changed. And the second area covered by the sensor 2 partially disappears. Therefore, in the calculation of the degree of coincidence of the surrounding environment information in the overlapping area of the surrounding environment recognition sensor 3, the coincidence degree recording unit 1e records that the degree of coincidence is low because there is no overlapping area.

  Subsequently, as shown in the example of FIG. 7, the sensor abnormality determination unit 1 h of the ECU 1 loads the degree of coincidence of the surrounding environment information stored by the degree of coincidence recording unit 1 e at the timing immediately before the collision from the memory of the ECU 1. Then, the sensor abnormality determination unit 1h determines the degree of coincidence of the surrounding environment information related to the vehicle 20 at the timing immediately before the loaded collision, and the vicinity recorded by the coincidence degree recording unit 1e at the timing immediately after the collision as shown in the example of FIG. Compare the degree of coincidence of environmental information. As a result of the comparison, the sensor abnormality determination unit 1h has a high degree of coincidence of the surrounding environment information at the timing immediately before the collision as shown in the example of FIG. 7, but the surrounding environment information at the timing immediately after the collision as shown in the example of FIG. When the degree of coincidence is low, the degree of coincidence between the two is not the same, so it is determined that an abnormality has occurred in at least one of the sensor 1 and the sensor 2 before and after the collision. This is because an abnormality such as an axis misalignment occurs in at least one of the sensor 1 and the sensor 2 due to the collision, and changes due to the collision in either the first area covered by the sensor 1 or the second area covered by the sensor 2. This is because it is considered that this occurred. As described above, after the collision detection unit 1g detects a collision, the sensor abnormality determination unit 1h determines whether the first sensor and the second sensor are not overlapped in the overlapping region. It is determined that there is an abnormality in at least one of them. Specifically, after the collision detection unit 1g detects a collision, the sensor abnormality determination unit 1h determines that the first sensor and the second sensor do not detect the same situation in the overlapping region. It is determined that at least one of the sensor and the second sensor is abnormal.

  Here, the sensor abnormality determination unit 1h also matches the surrounding environment information before and after the collision in the second overlapping area where the first area covered by the sensor 1 and the third area covered by the sensor 3 partially overlap. You may compare degrees. Thereby, the sensor abnormality determination part 1h can also determine whether abnormality has arisen in either the sensor 1 or the sensor 2 based on the comparison result of the coincidence degree of the surrounding environment information in the second overlapping region. . In the example of FIG. 9, the sensor abnormality determination unit 1h has a situation in which the first region and the second region do not overlap in the overlapping region, specifically, the sensor abnormality determination unit 1h When the two sensors do not detect the same situation, it is determined that at least one of the first sensor and the second sensor is abnormal. At this time, it is unknown whether an abnormality has occurred in the first sensor or an abnormality has occurred in the second sensor. Therefore, if the comparison result of the degree of coincidence of the surrounding environment information in the second overlapping region is approximately the same before and after the collision, the sensor abnormality determination unit 1h has no abnormality in the first sensor and has an abnormality in the second sensor. Determine that it has occurred. On the other hand, if the comparison result of the degree of coincidence of the surrounding environment information in the second overlapping region is not the same before and after the collision, there is no abnormality in the second sensor, and there is an abnormality in the first sensor. It is determined that an abnormality has occurred in both the first sensor and the second sensor.

  Returning to FIG. 1, the description of the configuration of the vehicle surrounding state estimation device according to the present invention will be continued. Of the ECU 1, the travel control unit 1i is based on various information indicating the vehicle momentum transmitted from the vehicle momentum detection sensor 2 and the situation around the host vehicle predicted by the surrounding situation prediction unit 1b. It is the travel control means which performs the travel control which controls. The travel control unit 1i is an area in which the vehicle 10 can travel, for example, indicated by the vehicle speed and acceleration of the vehicle 10 included in the information indicated by the vehicle momentum, and the predicted situation around the host vehicle (for example, the situation in the vehicle surrounding area). The vehicle 10 calculates a travel locus, a travel speed, and the like that the vehicle 10 can avoid the obstacle based on various information indicating the obstacle and the position of the obstacle to be avoided. Then, the traveling control unit 1 i outputs a control signal based on the calculation processing result to the actuator 4 and operates the actuator 4 to execute the traveling control. For example, the traveling control unit 1i controls the steering angle of the steered wheels of the vehicle 10 via the actuator 4 such as EPS, thereby performing steering assistance so that the vehicle 10 avoids an obstacle. The traveling control unit 1i may execute the steering assistance in combination with the brake assistance so that the obstacle can be avoided more reliably. In this way, the traveling control unit 1i functions as a traveling control unit that avoids the movement of the vehicle 10 to the position of the obstacle.

  Here, when it is determined by the collision avoidance determination unit 1f that a collision with an obstacle can be avoided, a collision with an obstacle can be avoided by performing the above-described traveling control by the processing of the traveling control unit 1i. It is. However, if the collision avoidance determination unit 1f determines that a collision with an obstacle cannot be avoided, the primary collision may not be avoided even by the processing of the traveling control unit 1i. Even in such a case, for safety reasons, the vehicle 10 is controlled immediately after the primary collision and moved to a safe place to minimize the impact (impact) caused by the secondary collision that may occur next. desirable. The driving control unit 1i can perform driving support control for avoiding secondary collision by performing traveling control even after the primary collision. In this case, the surrounding environment recognition sensor 3 is affected by the influence of the primary collision. It is also necessary to consider the possibility of abnormality.

  Therefore, in the present embodiment, based on the situation of the area where there is a possibility of a secondary collision estimated immediately before the collision, the situation around the vehicle immediately after the collision is predicted, and travel control for avoiding the secondary collision is performed. It is running. Further, in the present embodiment, traveling control for avoiding secondary collision is performed after predicting the situation around the vehicle immediately after the collision according to the sensor state (normal / abnormal) of the surrounding environment recognition sensor 3.

  Specifically, in this embodiment, the surrounding situation prediction unit 1b is a prediction area recorded in the surrounding situation recording unit 1d after the collision detection unit 1g detects a collision and before the collision detection unit 1g detects a collision. Based on the situation, the situation around the host vehicle is predicted. Further, the surrounding situation prediction unit 1b is set before the collision of the surrounding environment recognition sensor 3 (ambient situation detection means) that is determined to be abnormal by the sensor abnormality determination unit 1h after the collision detection unit 1g detects the collision. The situation of the abnormality recognition area in the vehicle surrounding area corresponding to the detected area is predicted based on the situation of the prediction area recorded by the surrounding situation recording section 1d before the collision detection section 1g detects a collision. In addition, the surrounding situation prediction unit 1b determines the surrounding environment recognition sensor for the situation of the normal recognition area in the vehicle surrounding area corresponding to the detection area of the surrounding environment recognition sensor 3 determined to be normal by the sensor abnormality determination unit 1h. 3 based on the detection result of 3. Then, after the collision detection unit 1g detects a collision, the traveling control unit 1i avoids the occurrence of a secondary collision around the host vehicle based on the situation in the vehicle surrounding area predicted by the surrounding situation prediction unit 1b. The travel control is performed.

  With reference to FIG. 10, the outline | summary of the prediction process of the secondary collision occurrence area | region in this embodiment is demonstrated. In FIG. 10, as the first collision timing, as shown in FIG. 4 described above, the vehicle 10 as the host vehicle overtakes another vehicle 30 that precedes the same traveling lane and travels in the opposite lane. The timing at which the vehicle 20 first collides is assumed.

  In FIG. 10, the surrounding situation estimation unit 1c is N seconds at a time point of M seconds immediately before the collision (that is, a time point when the collision avoidance determination unit 1f determines that it is impossible to avoid the primary collision with the vehicle 20). An area in which a subsequent peripheral obstacle (the vehicle 30 with the possibility of a secondary collision in FIG. 10) exists is estimated. Specifically, as shown in FIG. 5 described above, the surrounding state estimation unit 1c determines that the collision avoidance determination unit 1f cannot avoid a collision with the vehicle 20, and then the collision detection unit 1g detects that the vehicle 20 Based on the detection result of the surrounding environment recognition sensor 3 before detecting the collision with the vehicle 20, the situation of the prediction region where the occurrence of the secondary collision with the vehicle 30 around the own vehicle when the vehicle 20 collides is estimated. To do. At this timing, the surrounding situation recording unit 1d transmits the situation of the prediction region estimated by the surrounding situation estimation unit 1c in the memory of the ECU 1 and records it.

  In FIG. 10, the surrounding situation prediction unit 1b determines the situation of the vehicle surrounding area around the host vehicle at a time point of N seconds immediately after the collision (that is, when the collision detection unit 1g detects a primary collision with the vehicle 20). Predict. Specifically, immediately after the collision, the surrounding situation prediction unit 1b further uses various information indicating the vehicle momentum transmitted from the vehicle momentum detection sensor 2 in addition to the surrounding environment information acquired by the surrounding environment information acquisition unit 1a. Then, the position and inclination state of the host vehicle are specified. Subsequently, the surrounding situation prediction unit 1b applies the identified position and inclination of the host vehicle to the prediction area recorded in the memory of the ECU 1 at the point of M seconds immediately before the collision, thereby determining the collision position. Perform (primary prediction). As described above, when it is determined that the primary collision cannot be avoided by the primary prediction process of the surrounding situation prediction unit 1b, the region where the secondary collision occurs is determined using the ECU memory information recorded immediately before the collision. Can be predicted.

  Furthermore, at the time of N seconds immediately after the collision, the sensor abnormality determination unit 1h determines whether or not abnormality has occurred in each of the surrounding environment recognition sensors 3 mounted on the vehicle 10 due to the influence of the primary collision. Then, the surrounding environment prediction unit 1b determines the surrounding environment recognition sensor 3 determined to be abnormal based on the determination result of the sensor abnormality determination unit 1h (the sensor 2 covering the detection area on the right front side of the vehicle in FIG. 10). It is predicted that the area near the collision position cannot be normally recognized at N seconds immediately after the collision. Therefore, the surrounding state prediction unit 1b detects the collision before the collision of the surrounding environment recognition sensor 3 (sensor 2 in FIG. 10), which is determined to be abnormal by the sensor abnormality determination unit 1h after the collision detection unit 1g detects the collision. The situation of the abnormality recognition area in the vehicle surrounding area corresponding to the set detection area is predicted based on the situation of the prediction area recorded by the surrounding situation recording unit 1d at the point of M seconds immediately before the collision. That is, in FIG. 10, the situation of the abnormality recognition area is predicted based on the situation of the prediction area (part restored from the ECU memory in FIG. 10) recorded in the ECU memory corresponding to the range of the abnormality recognition area. .

  In addition, the surrounding situation prediction unit 1b determines that the surrounding environment recognition sensor 3 is determined to be normal based on the determination result of the sensor abnormality determination unit 1h (in FIG. 10, sensors 1 and 3 that cover areas other than the detection area on the right front of the vehicle). With respect to ˜6), it is predicted that each detection area can be normally recognized even at the point of N seconds immediately after the collision. Therefore, the surrounding situation prediction unit 1b is provided in the vehicle surrounding region corresponding to the detection region of the surrounding environment recognition sensor 3 (sensors 1 to 3 to 6 in FIG. 10) determined to be normal by the sensor abnormality determination unit 1h. The state of the normal recognition region is predicted based on the detection result of the surrounding environment recognition sensor 3. Subsequently, the surrounding situation prediction unit 1b predicts the situation of the abnormality recognition area predicted based on the ECU memory information recorded immediately before the collision and the situation of the normal recognition area predicted based on the surrounding environment information detected immediately after the collision. And the situation of the vehicle surrounding area at the point of N seconds immediately after the collision is predicted.

  As described above, the area covered by the sensor in which the abnormality has occurred after the primary collision is recorded immediately before the collision based on the determination result indicating the presence or absence of the sensor abnormality by the above-described processing as the secondary prediction of the surrounding situation prediction unit 1b. By utilizing the detected ECU memory information, the situation around the host vehicle can be predicted using the detection result actually detected at the timing immediately after the collision for the area covered by the normal sensor even after the primary collision. The prediction accuracy of the region where the next collision occurs can be further increased.

  Then, with reference to FIGS. 11-15, the various processes performed by the above-mentioned vehicle surrounding condition estimation apparatus are demonstrated. FIG. 11 is a flowchart showing an example of basic processing of the vehicle surrounding situation estimation apparatus according to the present invention. FIG. 12 is a flowchart illustrating an example of the surrounding situation estimation process immediately before the collision. FIG. 13 is a flowchart showing an example of the coincidence degree recording process immediately before the collision. FIG. 14 is a flowchart illustrating an example of sensor abnormality determination processing immediately after a collision. FIG. 15 is a flowchart illustrating an example of the surrounding situation prediction process according to the sensor state.

  As shown in FIG. 11, the surrounding environment information acquisition unit 1a receives and acquires surrounding environment information transmitted from the surrounding environment recognition sensor 3 and indicating the surrounding conditions of the vehicle such as a moving object and a stationary obstacle around the vehicle. (Step S10).

  The collision avoidance determination unit 1f is based on various information indicating the vehicle momentum transmitted from the vehicle momentum detection sensor 2 and the surrounding environment information acquired by the surrounding environment information acquisition unit 1a in step S10. It is determined whether or not a collision with an external object can be avoided (step S20). In the present embodiment, the collision avoidance determination unit 1f, for example, the relative position and relative speed between an object outside the vehicle indicated by the surrounding environment information and the vehicle 10, the vehicle speed and acceleration of the vehicle 10 included in the information indicated by the vehicle momentum, and the like. Based on the above, a time until a collision between an object outside the vehicle and the vehicle 10 (so-called collision prediction time (Time-To-Collision: TTC)) is calculated. Then, the collision avoidance determination unit 1f determines that collision can be avoided if the calculated TTC is equal to or greater than a predetermined threshold, and determines that collision avoidance is not possible if the calculated TTC is less than the predetermined threshold.

  In step S20, when the collision avoidance determination unit 1f determines that collision avoidance is impossible (step S20: Yes), the process proceeds to the next step S30, and when it is determined that collision avoidance is possible (step S20: No). ), The process returns to step S10.

  In step S20, when it is determined by the collision avoidance determination unit 1f that collision avoidance is impossible (step S20: Yes), the surrounding situation estimation unit 1c and the surrounding situation recording unit 1d perform estimation and recording of the surrounding situation at the time of the collision. (Step S30). The surrounding state estimation process immediately before the collision performed in step S30 in FIG. 11 will be described with reference to FIG.

  As shown in FIG. 12, the surrounding state estimation unit 1c estimates an area where there is an obstacle on the road after N seconds, which is the timing immediately after the collision (step S31). In step S31, the surrounding state estimation unit 1c is based on the surrounding environment information acquired by the surrounding environment information acquiring unit 1a before the collision detecting unit 1g detects a collision, as shown in FIGS. Thus, the situation of the prediction area where the occurrence of the secondary collision around the host vehicle at the time of the collision is predicted is estimated. Then, the surrounding situation recording unit 1d transmits the situation of the prediction region estimated in step S31 in association with the estimated time in the memory of the ECU 1 and records it (step S32). Thereafter, the process proceeds to step S40 in FIG.

  Returning to FIG. 11, the coincidence degree recording unit 1e performs a process of calculating and recording the coincidence degree of the surrounding environment information in the straddle region between the sensors at the timing immediately before the collision (step S40). Here, when it is determined by the collision avoidance determination unit 1f that collision avoidance is impossible (step S20: Yes), the time immediately before the collision is determined because the time TTC until the collision is determined to be less than the predetermined threshold. . The coincidence degree recording process performed in step S40 of FIG. 11 will be described with reference to FIG.

  As illustrated in FIG. 13, the coincidence degree recording unit 1 e includes the strength, brightness, color, relative position, and the like related to the recognition target (for example, an obstacle determined to be a collision avoidable) in the straddle region between the sensors. The degree of coincidence of the surrounding environment information is calculated (step S41). For example, as shown in FIG. 7 described above, when the coincidence degree recording unit 1e determines that collision avoidance is impossible by the collision avoidance determining unit 1f (that is, the timing immediately before the collision), the overlapping area of the surrounding environment recognition sensor 3 The degree of coincidence of the surrounding environment information is calculated. Then, the coincidence degree recording unit 1e transmits the degree of coincidence calculated in step S41 in association with the calculation time in the memory of the ECU 1 and records it (step S42). Thereafter, the process proceeds to step S50 in FIG.

  Returning to FIG. 11, in step S50, when the collision detection unit 1g detects a primary collision between the vehicle 10 and the obstacle (step S50: Yes), the process proceeds to the next step S60. When the primary collision with the object is not detected (step S50: No), the process is repeated from step S30 until the primary collision is detected. By repeating the processing from step S30, the surrounding situation recording unit 1d estimates the surrounding situation from the time point after the collision avoidance determining unit 1f determines that the collision avoidance is impossible until the collision detecting unit 1g detects the collision. The state of the prediction area estimated by the unit 1c is recorded.

  If the collision detection unit 1g detects a primary collision in step S50 (step S50: Yes), the surrounding situation prediction unit 1b records the surrounding situation prediction unit 1d before the collision detection unit 1g detects the primary collision. Based on the situation of the predicted region, the situation of the vehicle peripheral region around the host vehicle is predicted (step S60). For example, as shown in FIG. 10 described above, the surrounding situation prediction unit 1b performs N-second time immediately after the collision (that is, the time when the collision detection unit 1g detects the primary collision with the vehicle 20 as the primary prediction processing). ) To predict the situation of the area around the vehicle. Specifically, immediately after the collision, the surrounding situation prediction unit 1b further uses various information indicating the vehicle momentum transmitted from the vehicle momentum detection sensor 2 in addition to the surrounding environment information acquired by the surrounding environment information acquisition unit 1a. Then, the position and inclination state of the host vehicle are specified. Subsequently, the surrounding situation prediction unit 1b applies the identified position and inclination of the host vehicle to the prediction area recorded in the memory of the ECU 1 at the point of M seconds immediately before the collision, thereby determining the collision position. Do. Thereafter, the process proceeds to step S70.

  After the collision detection unit 1g detects a collision in step S50, the sensor abnormality determination unit 1h determines whether or not there are abnormalities in the surrounding environment recognition sensors 3 mounted on the host vehicle (step S70). In step S70, the sensor abnormality determination unit 1h is, for example, a first sensor that detects the state of the first region around the vehicle 10, and a region that is different from the first region and is a part of the first region. Whether there is an abnormality in the second sensor that detects the situation of the second region around the vehicle 10 that overlaps with the vehicle 10 is determined. The sensor abnormality determination process performed in step S70 of FIG. 11 will be described with reference to FIG.

  As shown in FIG. 14, when the collision detection unit 1g detects a collision in step S50 of FIG. 11 (step S50: Yes), the sensor abnormality determination unit 1h detects all surrounding environment recognition sensors mounted on the vehicle 10. 3, the recognition state is set to “abnormal” (step S71). Here, when the collision detection unit 1g detects a collision (step S50: Yes), it is the timing immediately after the collision. Therefore, in step S71 in FIG. 14, the coincidence degree recording unit 1e also performs processing for calculating and recording the coincidence degree of the surrounding environment information in the straddle region between the sensors at the timing immediately after the collision as described above. Subsequently, the sensor abnormality determination unit 1h loads the degree of coincidence of the surrounding environment information recorded by the coincidence degree recording unit 1e at the timing immediately before the collision in step S42 of FIG. 13 from the memory of the ECU 1 (step S72). And the sensor abnormality determination part 1h is the periphery recorded by the coincidence degree recording part 1e at the timing immediately after the collision in step S71 and the coincidence degree of the surrounding environment information regarding the vehicle 20 at the timing immediately before the collision loaded in step S72. The degree of coincidence of the environmental information is compared (step S73).

  Based on the comparison result obtained in step S73, the sensor abnormality determination unit 1h has the same degree of coincidence in the overlapping area between the sensors immediately before the collision and the degree of coincidence in the overlapping area between the sensors immediately after the collision. It is determined whether or not there is (step S74). Note that the determination process in step S74 is performed for each pair of sensors having a straddle region.

  If the sensor abnormality determination unit 1h determines that the comparison result of the coincidence degree is approximately the same in step S74 (step 74: Yes), the recognition state of the corresponding surrounding environment recognition sensor 3 is changed from “abnormal” to “normal”. Update to Thereafter, the process proceeds to step S80 in FIG. On the other hand, if it is determined in step S74 that the comparison result of the degree of coincidence is not comparable (step 74: No), the recognition state of the corresponding surrounding environment recognition sensor 3 is not “normal” but “abnormal” is updated. Leave. Thereafter, the process proceeds to step S80 in FIG.

  Returning to FIG. 11, the surrounding situation prediction unit 1b modifies the prediction result predicted primarily in step S60 according to the determination result of the sensor abnormality determination unit 1h, so that the situation of the vehicle surrounding area around the host vehicle is determined. Prediction is made (step S80). In step S80, the surrounding situation prediction unit 1b detects the collision before the collision of the surrounding environment recognition sensor 3 (ambient situation detection unit) that is determined to be abnormal by the sensor abnormality determination unit 1h. The situation of the abnormality recognition area in the vehicle surrounding area corresponding to the set detection area is predicted based on the situation of the prediction area recorded by the surrounding situation recording section 1d before the collision detection section 1g detects a collision. . In addition, the surrounding situation prediction unit 1b determines the surrounding environment recognition sensor for the situation of the normal recognition area in the vehicle surrounding area corresponding to the detection area of the surrounding environment recognition sensor 3 determined to be normal by the sensor abnormality determination unit 1h. 3 based on the detection result of 3. For example, as shown in FIG. 10 described above, the surrounding situation prediction unit 1b is recorded immediately before the collision based on the determination result of the sensor abnormality determination unit 1h at the time of N seconds immediately after the collision as the secondary prediction process. The vehicle at the time of N seconds immediately after the collision is obtained by combining the situation of the abnormality recognition area predicted based on the ECU memory information and the situation of the normal recognition area predicted based on the surrounding environment information detected immediately after the collision. Predict the situation in the surrounding area. The surrounding state prediction process according to the sensor state performed in step S80 of FIG. 11 will be described with reference to FIG.

  As shown in FIG. 15, the surrounding state prediction unit 1 b has a “normal” recognition state for each of the plurality of surrounding environment recognition sensors 3 mounted on the vehicle 10 based on the determination result of the sensor abnormality determination unit 1 h. It is determined whether or not there is (step S81).

  When the surrounding state prediction unit 1b determines that the recognition state of the target surrounding environment recognition sensor 3 is “normal” in step S81 (step S81: Yes), the surrounding environment recognition sensor determined to be “normal”. The normal recognition region in the vehicle peripheral region corresponding to the detection region 3 is set to use the detection result detected by the peripheral environment recognition sensor 3 at the timing immediately after the collision (step S82). Specifically, in step S82, the surrounding situation prediction unit 1b sets the value of the weighting factor K1 for using sensor information to 1 and sets the value of the weighting factor K2 for using ECU memory information to 0. Thereafter, the process proceeds to step S84.

  On the other hand, when the surrounding state prediction unit 1b determines that the recognition state of the target surrounding environment recognition sensor 3 is “abnormal” in step S81 (step S81: No), the surrounding environment determined to be “abnormal” About the abnormality recognition area | region in the vehicle periphery area | region corresponding to the detection area | region of the recognition sensor 3, it sets so that the information of the applicable area | region from the memory of ECU1 may be used (step S83). In particular. In step S83, the surrounding situation prediction unit 1b sets the value of the weighting factor K1 for using sensor information to 0 and sets the value of the weighting factor K2 for using ECU memory information to 1. Thereafter, the process proceeds to step S84.

  Then, the surrounding situation prediction unit 1b sets the weight of the sensor information use weighting factor K1 set in step S82 and step S83 to the following expression “vehicle surrounding area situation = K1 × sensor information + K2 × ECU memory information”, and the ECU. The value of the weight coefficient K2 for using memory information is input. Thereby, the surrounding situation prediction unit 1b predicts the situation of the abnormality recognition area predicted based on the ECU memory information recorded immediately before the collision and the situation of the normal recognition area predicted based on the surrounding environment information detected immediately after the collision. Is predicted as the situation of the vehicle peripheral area immediately after the collision (step S84). Thereafter, the process proceeds to step S80 in FIG.

  Returning to FIG. 11, the traveling control unit 1 i performs traveling control for avoiding the occurrence of the secondary collision around the host vehicle based on the situation of the vehicle surrounding area predicted in step S80 (step S90).

  As described above, even when the surrounding situation cannot be detected due to an abnormality in the sensor due to a collision, as shown in FIGS. Thus, the surrounding situation of the vehicle can be estimated appropriately. Here, FIG. 16 is a diagram illustrating an example of a scene in which the secondary collision position with respect to the moving object on the road is predicted. FIG. 17 is a diagram illustrating an example of a scene in which a secondary collision position with respect to a stationary object on a road is predicted.

  The left diagram of FIG. 16 shows a scene in which the vehicle 10 as the host vehicle overtakes another vehicle 30 that precedes the same traveling lane and first collides with another vehicle 20 traveling in the opposite lane. In FIG. 16, when the vehicle 10 passes the vehicle 30, it is assumed that an abnormality has occurred in the surrounding environment recognition sensor 3 (for example, the sensor 2) mounted in the right front due to a collision in the right front with the vehicle 20. In such a situation, in the present embodiment, the sensor abnormality determination unit 1h determines that the recognition state of the surrounding environment recognition sensor 3 (for example, the sensor 2) mounted on the right front is abnormal. Moreover, it determines with the recognition state of the surrounding environment recognition sensor 3 (for example, sensor 1, 3-6) mounted in positions other than the right front by the process of the sensor abnormality determination part 1h being normal. Subsequently, as shown in the right diagram of FIG. 16, as the primary prediction result of the secondary collision position, the vehicle after the collision with the vehicle 20 is obtained from information such as the yaw rate and acceleration by the processing of the surrounding situation prediction unit 1 b. The position and inclination of 10 are predicted, and the approach position of the rear vehicle to be subjected to the secondary collision with the vehicle 30 is predicted from the ECU memory information recorded immediately before the collision. Further, as a secondary prediction result of the secondary collision position, the sensor that covers the left rear determined to be “normal” (for example, the sensor 5) by the processing of the surrounding situation prediction unit 1b is switched to the actual sensing information, The actual distance to the vehicle 30 is determined.

  The left diagram in FIG. 17 shows a scene in which the vehicle 40 primarily collides with the left front of the vehicle 10 while the vehicle 10 as the host vehicle is traveling on a lane having a flat wall on the left side and an uneven wall on the right side. Show. In FIG. 17, it is assumed that the vehicle 10 has an abnormality in the surrounding environment recognition sensor 3 (for example, the sensor 3) mounted on the left front when the vehicle 10 collides. In such a situation, in the present embodiment, the processing of the sensor abnormality determination unit 1h determines that the recognition state of the surrounding environment recognition sensor 3 (for example, the sensor 3) mounted on the left front is abnormal. Moreover, it determines with the recognition state of the surrounding environment recognition sensor 3 (for example, sensors 1-2, 4-6) mounted in positions other than the left front by the process of the sensor abnormality determination part 1h being normal. Subsequently, as shown in the right diagram of FIG. 17, as a primary prediction result of the secondary collision position, the host vehicle after the collision with the vehicle 40 is obtained from information such as the yaw rate and acceleration by the processing of the surrounding situation prediction unit 1 b. The position and inclination of 10 are predicted, and the distance from the uneven wall on the right side that is the target of the secondary collision is predicted from the ECU memory information including the road width information recorded immediately before the collision. Further, as a secondary prediction result of the secondary collision position, the sensor 2 covering the right front determined to be “normal” is switched to the actual sensing information by the processing of the surrounding situation prediction unit 1b, and the right uneven wall is The actual distance (d_right in FIG. 17) is determined. Also, the actual distance to the left flat wall (d_left in FIG. 17) is determined together with the road width information.

  16 and 17, when comparing the primary prediction result of the secondary collision position with the secondary prediction result of the secondary collision position, the primary collision out of the regions (i) to (iv) where there is a possibility of collision. The region (i) indicated by the prediction result and the region (iii) indicated by the secondary prediction result have different positions. This is because the area covered by the sensor determined to be normal after the collision is preferably predicted by switching to actual sensing information rather than using the ECU memory information recorded immediately before the collision. This shows that the situation can be predicted. As described above, according to the present embodiment, it is possible to appropriately estimate the surrounding situation of the vehicle even when the abnormality occurs in the sensor due to the collision and the surrounding situation cannot be detected.

  In the above-described embodiment, the coincidence degree recording unit 1e has a timing immediately before the collision (that is, a timing at which the collision avoidance determining unit 1f determines that collision cannot be avoided) or a timing immediately after the collision (that is, the collision detecting unit 1g). In the example described above, the degree of coincidence of the surrounding environment information in the overlapping region of the surrounding environment recognition sensor 3 is calculated and recorded at the timing when the collision is detected. The degree of coincidence of the surrounding environment information in the overlapping region of the surrounding environment recognition sensor 3 may be calculated and recorded (for example, when the engine is started or when the steering assist switch is turned on). Thereby, the sensor abnormality determination unit 1h confirms that there is no abnormality in the sensor mounted on the vehicle 10 before performing the process of comparing the coincidence degree of the overlapping areas between the sensors immediately before and after the collision. Can do.

  In the above-described embodiment, an example in which the travel control unit 1i performs travel control (for example, steering control or brake control) to avoid the occurrence of a secondary collision around the host vehicle has been described. The processing content of 1i is not limited to this. For example, in the present embodiment, the position of the white line can be estimated as the situation of the prediction area around the host vehicle when the vehicle collides, by the process of the surrounding situation estimation unit 1c. Thereby, even after the collision is detected by the processing in the surrounding situation prediction unit 1b, the travel defined by the white line around the host vehicle based on the situation of the predicted area including the position of the white line recorded before the collision is detected. Possible areas can be predicted. As a result, it is also possible to perform travel control (so-called lane keeping assist: LKA) for controlling the behavior of the host vehicle so as to follow the white line even after the collision is detected by the travel control unit 1i. As described above, according to the vehicle surrounding state estimation device of the present embodiment, it is possible to predict the travelable region that is preferably defined by the white line even after the primary collision. Even when it becomes impossible to detect, LKA control can be executed. Therefore, by executing the LKA control even after the primary collision, it is possible to reduce the situation where the host vehicle runs out of the lane that the vehicle is traveling after the collision. The occurrence of secondary collision with obstacles installed in the can also be reduced.

1 ECU (vehicle surrounding situation estimation device)
DESCRIPTION OF SYMBOLS 1a Ambient environment information acquisition part 1b Ambient condition prediction part 1c Ambient condition estimation part 1d Ambient condition recording part 1e A coincidence degree recording part 1f A collision avoidance determination part 1g A collision detection part 1h A sensor abnormality determination part 1i A travel control part 2 A vehicle momentum detection sensor 2a Acceleration sensor 2b Yaw rate sensor 2c Vehicle speed sensor 3 Surrounding environment recognition sensor 3a Sensor 1 (first sensor)
3b Sensor 2 (second sensor)
3c sensor 3 (third sensor)
4 Actuator

Claims (4)

Collision detection means for detecting that the host vehicle has collided with an object outside the vehicle;
Surrounding situation detection means for detecting the situation of the detection area around the vehicle,
Based on the detection result of the surrounding state detection unit before the collision detection unit detects a collision, a surrounding state estimation unit that estimates the state of the prediction region around the host vehicle when a collision occurs;
A surrounding situation recording means for recording the situation of the prediction area estimated by the surrounding situation estimation means;
After the collision detection means detects a collision, the situation of the vehicle surrounding area around the host vehicle is determined based on the situation of the predicted area recorded by the surrounding situation recording means before the collision detection means detects a collision. A surrounding situation prediction means to predict;
A vehicle surrounding state estimation device comprising: A collision avoidance determining means for determining whether or not a collision between the host vehicle and the object outside the vehicle detected by the surrounding state detecting means can be avoided;
The surrounding situation estimation means includes
After the collision avoidance determining means determines that collision avoidance is impossible, the situation of the prediction area is estimated,
The surrounding situation recording means is
The situation of the prediction region estimated by the surrounding situation estimation unit from a time point after the collision avoidance determination unit determines that collision avoidance is impossible to a point before the collision detection unit detects a collision. The vehicle surrounding situation estimation apparatus according to 1. After the collision detection unit detects a collision, a sensor abnormality determination unit that determines whether or not there is an abnormality in the surrounding state detection unit mounted on the host vehicle,
Further comprising
The surrounding situation prediction means includes
After the collision detection means detects a collision, the abnormality recognition area in the vehicle peripheral area corresponding to the detection area set before the collision of the peripheral situation detection means determined to be abnormal by the sensor abnormality determination means The situation is predicted based on the situation of the prediction area recorded by the surrounding situation recording means before the collision detection means detects a collision, and the sensor abnormality determination means determines that the situation is normal. The situation of the normal recognition area in the vehicle surrounding area corresponding to the detection area of the surrounding situation detecting means is predicted based on the detection result of the surrounding situation detecting means, and the situation around the own vehicle is predicted. The vehicle surrounding state estimation device according to 1 or 2. Travel control means for performing travel control for controlling the behavior of the host vehicle based on the situation of the vehicle surrounding area predicted by the surrounding situation prediction means after the collision detection means detects a collision;
The vehicle surrounding state estimation device according to any one of claims 1 to 3, further comprising:

Images